[French]

CASSOS

Computer-assisted Sector Opening Schedules

[Image IHM]

Click here to see our ON-LINE DEMONSTRATOR
  • 1999 : research (at the Global Optimization Lab ENAC/CENA) on the optimal airspace partitionning, using the sector capacities declared by the ATC centers
  • 2002-2007 :suspended
  • 2007 : research (at DSNA/DTI/R&D/POM) on airspace configuration forecasts, using a workload model based on air traffic complexity indicators and a neural network (see S2D2)
  • 2008-2010 : participatory design (with ATC and HMI experts) of a human-machine interface for airspace configuration forecasts
  • 2009-2011 : development of a mock-up HMI, contribution to the SESAR european program:
    • SESAR 4.7.1 Complexity management in en-route
    • SESAR 10.8.1 Complexity assessment and resolution
    • and marginally SESAR 7.5.4 Dynamic airspace configuration
  • End of 2011 : the development of CASSOS software at DSNA/DTI is interrupted, due to a reorganization and to the departure of the project's progammers and researchers.

This page presents a summary of the CASSOS project. If you really love details, you can also have a look at the former web page.

Problem statement

The airspace is currently divided into elementary sectors (or airspace sectors, or modules) that may be merged (or collapsed) into larger sectors, and operated as control sectors assigned to controller working positions. The way elementary sectors are combined into control sectors depends (at least in France) on the traffic flow through the airspace and on the workload of the controllers operating the sectors.

In operations, it is currently quite difficult to anticipate which control sectors might become overloaded in the next few hours, or the next day, even with an accurate traffic prediction. The problem is twofold. First, it is not easy to quantify the workload of a controller operating a given sector with a given traffic. Second, we don't know a priori which control sectors will actually be operated, because the airspace is dynamically partitionned into control sectors, depending on the controllers' workload.

Objective

Our main objective is to make realistic predictions of the airspace configuration. In the problem being addressed here, the traffic is not considered as a variable that can be adjusted, but as input data. So the traffic demand is not regulated by assigning departure slots, but instead the elementary sectors are combined as well as possible so as to build the most adapted airspace configuration, taking into account various operationnal constraints.

R&D approach

Research on the algorithms (David Gianazza)

We search optimal partitions of the airspace, while satisfying a number of constraints. The cost function being minimized depends on the controller workload of each control sector in the airspace partition. One of the difficulties is to estimate correctly this controller workload Thus, the research phase focused on two issues:

  • the workload prediction, using air traffic complexity indicators (see S2D2, in collaboration with Kévin Guittet) and a neural network,
  • the optimal partitionning of the airspace into control sectors.

algo_partitionnement

Participatory design of the HMI (Christophe Hurter)

The design of a human-machine interface (HMI) for the airspace configuration forecasts has been the subject of participatory design sessions involving several ATC (Air Traffic Control) experts and HMI experts.

Participants:

  • HMI designers: Christophe Hurter, Benjamin Tissoires, Gilles Tabard
  • ATC (Air Traffic Control) experts: Claude Chamayou, Géraldine Beboux, Florence Cressent
  • Researcher (algorithms): David Gianazza

[Image IHM]

HMI conception and development (Nicolas Saporito)

An iterative process, cycling through programming phases and participatory design sessions, allowed us to implement an HMI prototype (in Flex language). The current version features dynamic interactions, although only on static data for the time being.

[Image IHM]

Contributions to R&D programs, and industrial collaborations (Nicolas Saporito, David Szymanski)

The work presented here is part of the DSNA (Direction des Services de la Navigation Aérienne) contribution to SESAR, the major european R&D program for Air Traffic Management. An industrial partnership has also been initiated with Thalès ATM in order to develop a pre-industrial prototype (DECOMPLEX).

Pending issues

In the current version of the algorithms, the successive airspace partitions are completely independent one from the other. There remains to take into account the sector merge/split operational constraints between successive partitions.

Testing the CASSOS software in an operational context would require to feed the algorithms with actual trajectory predictions.

Software

The CASSOS algorithms are written in Objective Caml, and the HMI prototype in Flex. The code is not public domain.

A library of neural networks was implemented in Ocaml and may be soon available under L-GPL license (see the software page).

Main publications

Last Update: